Locking the Fridge: Food, Choice and Restrictive Practice

Locking the fridge: diabetes, food and the restrictive practice nobody authorises

The NDIS Quality and Safeguards Commission’s Regulated Restrictive Practices Guide defines environmental restraint as a practice that “restricts a person’s free access to all parts of their environment, including items or activities.”

Among the examples it gives:

“Locking a door, cupboard or fridge to prevent a person’s access.”

“Preventing access to a certain area in their environment, such as the backyard or kitchen.”

“Preventing or placing restrictions on a person engaging in an activity such as cooking.”

Now consider what managing type 2 diabetes in supported accommodation actually involves. It involves influencing what an adult eats, in a house where somebody else does the shopping, the cooking and the portioning.

Those two things collide, constantly, in houses all over Australia. And the collision is resolved in one of two ways, both of which are failures.

Failure one: the fridge gets locked and nobody tells anybody. Cupboards keyed. Snacks held in the office. The person’s own money managed so they cannot buy food. A worker sent with them to the shops to stop them buying things. All of it done for the person’s diabetes, all of it well-intentioned, none of it authorised, reported, or written into a behaviour support plan. It is a regulated restrictive practice operating as a house rule.

Failure two: the provider has heard that food restriction is a restrictive practice and therefore provides no structure at all. Unlimited access to whatever is in the house. No meal planning. No dietitian. No conversation. HbA1c running at 11 or 12 per cent for years, and a foot amputated at 54. The provider has complied scrupulously with a rights framework and comprehensively failed its duty of care.

Neither is the answer, and there is a defensible path between them. This guide is about finding it.

We support NDIS participants across New South Wales, including people with diabetes living in supported accommodation. We have an obvious stake in this question, so we have tried to write it straight — including the parts where our own sector’s instincts are wrong in both directions.

What the framework actually says

The detail matters here, so it is worth being precise.

Environmental restraint is one of the five regulated restrictive practices under the NDIS framework, alongside seclusion, chemical restraint, mechanical restraint and physical restraint. Where a regulated restrictive practice is used, the provider must follow the authorisation, consent and reporting requirements of the relevant state or territory, must have the practice clearly identified in a behaviour support plan prepared by an NDIS behaviour support practitioner, must keep records, and must report its use to the Commission. In practice that means an interim behaviour support plan within one month, a comprehensive plan within six, and monthly reporting of use.

And here is the qualifier that does most of the work. The guide frames regulated restrictive practices as practices “beyond ordinary community standards” used to address a behaviour of concern that can cause harm.

Its own illustration of the boundary is useful: locking the front door at night for safety is not environmental restraint, because the person can still leave if they choose. The lock is not there to control them.

So the question in any given situation is not simply “is there a lock?” It is: what is this for, and whose behaviour is it directed at?

A related clarification from the same guide, worth knowing because it reassures people unnecessarily worried: locked medication is not environmental restraint and does not need reporting as a restrictive practice. Providers administering medication are required to store it safely under the NDIS Practice Standards. So a locked medication cupboard containing insulin is an ordinary obligation, not a restriction.

The distinction that resolves most of this

Restriction removes an option. Support changes which options are easy, informed and attractive.

That sentence is the practical test, and applying it clears up most situations.

Things that are not restrictive practices

What the house buys. Nobody has a right to have any particular food in their house. Deciding that the weekly shop includes less soft drink and more of something else is a purchasing decision, and purchasing decisions are made in every household on earth. If the Coke is not bought, there is no Coke to lock away.

This is the single most useful realisation in this area, and it is the one providers most often miss while reaching for a key. The question “how do we stop him drinking six cans a day” has a different and much better answer than a locked fridge: do not buy six cans a day. That is not a restriction on the person; it is a decision about a shopping list, which somebody was always going to make.

The obvious caveat: it has to be done with the person, not to them. A shopping list imposed without discussion is a different thing from one worked out together. And where the person buys their own food with their own money, their choices are their own.

Portioning meals as served. Serving a meal of a particular size is ordinary meal service. Every household in the country does it. It becomes something else only if the person’s access to more food is then physically prevented.

Meal planning, menus and routines. Having a weekly menu is ordinary household practice, not a restriction.

Offering, prompting, encouraging, educating, discussing. All support. None of it restricts anything.

Presenting the easier option. Fruit in a bowl on the bench, water in the fridge, the biscuits in the back of the cupboard rather than on the table. This is how everybody manages their own eating and it restricts nothing.

Cooking with the person. Which is also a legitimate NDIS capacity-building goal, and produces far better outcomes than any amount of rule-setting.

Things that are restrictive practices

Locking the fridge, freezer, pantry or kitchen to prevent the person’s access.

Preventing the person entering the kitchen, or restricting them from cooking.

Physically preventing the person from taking food that is there.

Controlling the person’s own money so they cannot buy food — which raises its own separate and serious questions about financial autonomy.

Following or supervising the person specifically to prevent them obtaining food.

If any of those is happening, it needs to be named, authorised, planned and reported — not operated as a house practice.

The thing that has to be said about rights

An adult with capacity has the right to eat what they want.

Including badly. Including in ways that will harm them. Including after being given full information about the consequences. People without disabilities do this every day, and nobody locks their fridge.

Supported decision-making means supporting the person to make the decision — not making it for them. The starting point is the presumption of capacity. Capacity is decision-specific: somebody who needs help managing money may be entirely capable of deciding what they eat. And needing information in an accessible format is not the same as lacking capacity.

So for a person with capacity, the correct response to a diet that is worsening their diabetes is:

• Information they can actually use, in Easy Read format, more than once
• An honest conversation about what is happening to their body and what it will cost them — which is a conversation people are frequently not given, because it feels unkind
• Involvement in shopping, planning and cooking
• Making the better option the easy option
• A dietitian who talks to the person rather than about them
• Respecting the choice they then make, while continuing to offer

That last clause is the hard part. A provider that finds this uncomfortable should notice that the discomfort is not a reason to override somebody.

And the part that gets forgotten entirely. For a person whose life contains very few choices — who did not choose their house, their housemates, their workers, their roster or their activities — food may be one of the only things they do choose. Taking it away without putting anything in its place is not a clinical intervention. It is removing one of the last areas of autonomy somebody has, and it should be weighed as such.

If a plan to improve somebody’s diet does not include anything being added to their life, it is probably the wrong plan.

Where it genuinely gets difficult

It would be dishonest to leave it there, because there are situations where the rights answer and the duty of care answer genuinely conflict.

Where the person does not have capacity for this decision, even with full support, and unrestricted access is causing serious, documented harm.

Where food-seeking is a behaviour of concern — taking other residents’ food, eating non-food items, or eating to the point of medical crisis.

Prader-Willi syndrome, which is the clearest case and deserves naming specifically. Hyperphagia — persistent, severe hunger that is not relieved by eating — is a core feature of the syndrome, not a behaviour the person is choosing. Unrestricted food access in Prader-Willi syndrome is genuinely dangerous: it leads to rapid severe obesity, type 2 diabetes, and in some cases life-threatening acute events. Food security measures, including secured food storage and supervised access, are standard and internationally accepted management.

And in the NDIS context, that is still a regulated restrictive practice. The point is not that it should not happen. The point is that it should be properly authorised, planned, documented and reviewed — which is exactly what the framework is for. A house managing Prader-Willi syndrome with locked food storage and no authorisation is doing the right clinical thing in the wrong regulatory way, and is one complaint away from a serious problem.

The shared house problem. One resident needs food access restricted; the others do not. Locking the kitchen restricts four people to manage one.

The Commission’s guide illustrates the approach to this with a case study about a locked cutlery drawer in a shared kitchen, where the other participant was given a key to reduce the impact on their rights. The principle is the same here: the least restrictive option that achieves the purpose, with the impact on other residents actively minimised. A lockable fridge drawer rather than the whole fridge. A secured pantry rather than the whole kitchen. Keys held by residents who do not need the restriction. Separate food storage.

What is not acceptable is four adults losing access to their own kitchen because it is administratively simpler.

Doing it properly: the sequence

If you are a provider, a family member or a coordinator facing this, this is the order.

1. Get a clinical opinion in writing. A letter from the GP, endocrinologist or diabetes educator stating the clinical risk, the HbA1c, and what is at stake. This document is what makes everything downstream defensible, and without it you are implementing a house opinion.

2. Get a dietitian involved. This changes the character of the whole exercise. A dietitian’s plan is a clinical plan. A worker’s view about what the person should eat is not. A dietitian for diabetes sits on the health side, funded via a GP’s chronic condition management arrangements.

3. Establish capacity for this specific decision, properly, with support. Not an assumption either way.

4. Exhaust supported decision-making and non-restrictive options first, and write down that you did. Easy Read information, repeated conversations, involvement in shopping and cooking, changing what is bought, changing what is presented and where, adding alternatives rather than only removing things. The documentation of what was tried is as important as the trying, because it is what demonstrates a least-restrictive approach.

5. If restriction is genuinely needed, do it through the framework. Involve an NDIS behaviour support practitioner. Obtain authorisation under the relevant state or territory process. Get it into a behaviour support plan with the practice named. Report its use to the Commission. Include a reduction strategy with actual steps, not an aspiration.

6. Minimise the scope. The smallest restriction that achieves the purpose, affecting the fewest people, for the shortest time.

7. Review it, genuinely. A restriction that has been in place for three years without review is not being managed; it is being maintained.

8. Add something. If autonomy is being reduced in one area, increase it somewhere else. This is not a soft gesture — it is the difference between a clinical plan and a punishment.

What should be in writing

• The clinical risk, from a clinician, with the numbers
• A dietitian’s plan
• What capacity assessment was done, for this decision, and what it concluded
• What supported decision-making was provided, in what format, how many times
• What non-restrictive strategies were tried and what happened
• The restriction itself, precisely described — what, when, who holds the key, what the person can access freely
• The authorisation
• The behaviour support plan naming the practice
• The reduction plan with dates
• The impact on other residents and how it has been minimised
• The review date
• What has been added to the person’s life alongside what has been restricted

A note to providers

Two uncomfortable observations.

The fridge is often locked because it is easier than the alternative. Meal planning with a person, cooking with them, shopping with them, having repeated honest conversations and accepting that they may still choose badly — all of that takes more worker time than a key does. When a restriction is chosen over support, it is worth asking honestly whether the reason is clinical or operational.

And the reverse. A provider that lets somebody’s diabetes run uncontrolled for a decade because it was frightened of the restrictive practice framework has not protected that person’s rights. It has protected itself. The framework does not prohibit restriction; it requires authorisation and oversight. A provider that will not seek authorisation because the process is onerous is making a different choice than it thinks it is.

The honest position is that this is difficult, that both errors are common, and that the way through is clinical input, documented supported decision-making, and the willingness to use the authorisation process when it is genuinely needed.